Forward Deployed Engineers Are Just Professional Services

0:00 - It's professional services, broadly with lipstick on.

0:03 You know, I would say 22-year-old Amy

0:06 did not have the CRO aspiration at the time.

0:09 I think luckily for my career, I started off at IBM.

0:13 - My guest today is Amy Meni,

0:15 Chief Revenue Officer at Sourcegraph,

0:18 the code intelligence platform that giant engineering orgs

0:21 use to make sense of decades of code.

0:24 Amy's path to the revenue seat

0:26 is one of the most unusual I have come across.

0:29 She studied computer science, started her career

0:32 as a developer, and rose through Solutions Architecture

0:35 and Professional Services before taking on all of revenue.

0:39 - It feels good to win that deal.

0:40 It feels good to launch that customer.

0:42 It feels good to, you know, get to the end

0:44 of this work stream or help them, you know,

0:47 be able to do something that they weren't able to do before.

0:50 Like there is an immense amount of satisfaction

0:55 that I get right from those.

0:57 Like, you know, the adrenaline hit of that.

0:59 Like that FDE-type package is more tied

1:02 to like an outcome that's being delivered.

1:04 And we're agreeing to some dollar amount.

1:07 They're not thinking about that third party.

1:09 They're thinking about, man, the software doesn't work for me.

1:12 You know, I think that maturity curve, right, is very real.

1:15 Learn to do it well, learn to do it consistently first.

1:20 And then you kind of earn the right

1:21 to start to charge for it.

1:27 (logo whooshing)

1:29 - Amy, you studied computer science.

1:31 You started out as a developer.

1:33 Now you run all of revenue at Sourcegraph.

1:36 Walk me through how a developer

1:38 ends up as a chief revenue officer.

1:40 - Yeah, yeah, happy to.

1:41 Thanks for having me.

1:43 You know, I would say 22-year-old Amy

1:46 did not have the CRO aspiration at the time.

1:49 Coming out of college, I was absolutely certain

1:53 I was going to do development,

1:54 kind of path towards, you know, CTO/CIO

1:57 was what I thought was on the horizon for me.

2:01 I think luckily for my career,

2:03 I started off at IBM massive organization

2:07 pretty much immediately threw me on to a project

2:10 with the US Navy.

2:12 And, you know, looking back,

2:15 thankfully that was the first project.

2:17 I was pretty bored through it.

2:20 You know, working on these like very large,

2:23 massive pieces of software, very waterfall.

2:27 You know, you're releasing quarterly, right?

2:30 I was sitting behind a computer and I was like,

2:31 this is just not for me.

2:34 Luckily being at IBM, where they do so many things,

2:37 I was able to move from that really

2:39 into technical consulting.

2:43 You know, IBM has a habit of acquiring software, right?

2:47 Putting their name on it

2:49 and kind of taking it forward from there.

2:51 And so right time, right place for me,

2:53 they had just purchased an asset management software

2:56 called Maximo.

2:57 And they were looking for people with technical chops

3:00 that could learn that and go into the field

3:03 and help, you know, sell that

3:05 and implement it for customers.

3:07 And so it was like pretty early on in my career

3:09 that I made the pivot

3:11 that was more directly customer facing.

3:15 And it really took off from there.

3:16 I spent the first part of my career in consulting.

3:18 I loved it.

3:19 I felt like I got a ton of exposure

3:21 to different industries and technologies.

3:25 I feel like I learned a lot in a short amount of time.

3:29 You're working on RFPs.

3:30 You're putting together proposals.

3:32 Granted, I was doing more of the solutioning aspect of those,

3:35 but you know, you're having to kind of do the full suite

3:38 of things when you're in a consultant type role.

3:42 And so I feel like,

3:43 especially that consultant part of my career,

3:45 I was at this interesting intersection of, you know,

3:48 of course being very customer facing,

3:52 adjacent to sales and working with really,

3:56 really technical products.

3:59 Asset management, right?

4:01 That's like a very complicated comp.

4:03 It's like on the level of ERP systems type complexity.

4:08 And so, you know, fast forward

4:10 to make the decision to leave consulting.

4:12 I felt like I got all of this exposure,

4:14 but it felt like a mile wide an inch deep.

4:17 And so when I left consulting

4:19 and really wanted to focus more vertically,

4:23 I started off in travel tech.

4:24 This was actually where Solution Architecture

4:27 Professional Services,

4:30 company by the name of SwitchFly.

4:32 We had kind of the full suite of capabilities.

4:35 We were doing B2B and B2B2C software.

4:38 So like white label behind the scenes.

4:40 You don't even know that you're interacting

4:41 with the software and you are.

4:43 There was a big professional services portion of that

4:46 that I was doing, working with customers,

4:48 trying to understand what it was that they need,

4:50 actually scoping it out, building SOWs,

4:54 working with internal teams to kind of deliver

4:56 that functionality that they needed.

4:58 Honestly, it sounds a lot like FDE

4:59 that I know we'll come back to.

5:02 And, you know, building new capabilities

5:05 and features specific to those customers.

5:07 Took that experience to segment, as you mentioned,

5:11 Solution Architecture team joined them

5:13 about 50 million in revenue.

5:14 We were doing a lot and kind of hit this moment

5:17 where we were trying to figure out

5:18 kind of repeatable playbooks for ourselves.

5:22 Turned that into a professional services organization

5:25 as well, stayed there through the Twilio acquisition.

5:29 And then that landed me at Sourcegraph.

5:32 I was, you know, a lot of it is my passion

5:35 and kind of background as you pointed out

5:37 from the development side of things,

5:40 but it's an extremely technical product.

5:42 We work with the most sophisticated

5:43 engineering organizations in the world.

5:45 And so I feel like so much of my past history

5:48 has led me to this moment.

5:50 I joined on the technical success team.

5:53 We were a small team of six

5:56 at the time when I joined right alongside sales.

5:59 So really this like, you know,

6:01 getting closer and closer to directly revenue,

6:05 it was very organic throughout, you know,

6:08 the entire five and a half years

6:10 that I've been at Sourcegraph,

6:11 I've been right here alongside.

6:13 Similarly, like you're hearing, you know,

6:15 stood up professional services,

6:17 started to sell that, scope that.

6:19 But being the technical product that we are

6:21 and one of our amazing competencies as a company

6:25 is our ability to land and really grow our customers.

6:29 I think that lends itself

6:30 to a lot of my background and experience.

6:33 And so, you know, it is a non-standard

6:39 kind of career path from what you tend to see.

6:42 But I think for me and my journey,

6:44 it actually feels like everything has led to this moment

6:48 in a very weird way.

6:50 - Was there a flip that switched

6:53 and something that changed

6:54 the moment you were doing something for a customer

6:58 versus just developing something behind the scenes?

7:02 - I almost feel like it was an immediate shift for me.

7:06 So I mentioned, you know, at IBM,

7:09 I get put onto this new technology that really,

7:13 very few people at IBM knew how to use.

7:16 And, you know, hungry scrappy, you know,

7:19 22 year old at the time learning this new technology,

7:22 getting to then immediately be deployed out, you know,

7:26 I'm onsite five days a week, you know,

7:29 in these customer environments

7:31 trying to get the software up and running for them.

7:33 I loved it immediately.

7:35 The problem solving, the figuring out what they need,

7:39 you know, I'm working with a gas utility on this project.

7:43 I'm working with a steel company on this project.

7:45 I'm working with telecommunications company

7:47 on this project, right?

7:48 So different.

7:51 And I think just instantaneously

7:53 that ability to interact with people,

7:56 understand what they need, what they're trying to do,

7:58 what they have in place, where we're trying to go.

8:01 And of course applying, you know, my technical skills

8:04 and actually like making the software work for them.

8:07 It was immediately just this instant

8:10 kind of trifecta for me.

8:13 - Yeah, I've always felt too,

8:15 there's such a difference between doing something internal

8:18 versus doing something external.

8:20 There's like a level of polish.

8:22 And I mean this in a good way,

8:24 a level of adrenaline too, of like,

8:27 oh, I need to be on my A game.

8:28 Like this matters.

8:30 And if I do well, this is a win.

8:33 And I just think it's totally different.

8:35 For anyone who has only done like development or ops,

8:38 like behind the scenes only worked internal,

8:41 like there's just something different

8:42 about doing it for a customer.

8:43 - Yeah, yeah.

8:45 You know, there's, I think you said discomfort.

8:48 There's a stress to it, but it's a good stress.

8:51 You know, athlete my whole life, right?

8:54 Team, you know, big into team sports.

8:56 And like, you kind of get a rush of that

8:59 in a professional environment.

9:02 It feels good to win that deal.

9:03 It feels good to launch that customer.

9:04 It feels good to get to the end of this work stream

9:07 or help them, you know, be able to do something

9:11 that they weren't able to do before.

9:13 Like there is an immense amount of satisfaction

9:18 that I get right from those,

9:19 like, you know, the adrenaline hit of that.

9:23 Yeah.

9:23 - 100%.

9:24 Yeah, I played a number of sports.

9:26 Main one I did was wrestling, middle school, high school.

9:31 And there's such a difference between practice

9:33 and game day.

9:35 And that level of speed and that level of attention

9:39 that you have and adrenaline in that moment is just,

9:43 it's pretty tough to replicate.

9:44 And I think in a business context,

9:46 game day is when you're doing something in front

9:49 of a customer, a prospect.

9:52 And that's when I think some of the best work is done too.

9:55 - Yeah, I agree with that completely.

9:59 So walk me through this craze of forward deployed engineer.

10:06 I know everybody's throwing it out.

10:08 Everybody's coming up with their own version too,

10:10 like forward deployed XYZ.

10:14 Is this just the same thing that we've always been doing?

10:16 Just call it something different?

10:17 Why do you think people are hanging onto this so much?

10:20 - Yeah, I think craze is definitely the right word there.

10:24 It's so interesting to see this evolve.

10:29 The original co-founders of Sourcegraph,

10:32 they themselves were forward deployed engineers

10:34 at Palantir earlier in their careers.

10:37 - Part of Genesis of the term.

10:39 - Yeah, absolutely.

10:41 And so it's so interesting to see the evolution of it.

10:45 I actually do believe that it's professional services

10:50 broadly with lipstick on,

10:52 but you're seeing it evolve in different ways.

10:56 The Palantir model I would say is like,

10:58 I'm going into the environment,

11:00 I'm building them whatever they want.

11:03 It's completely custom, bespoke to what they need.

11:05 That's my priority is what is this customer

11:09 trying to achieve?

11:11 And I think that that certainly still exists.

11:13 I mean, that's the work that I was doing at Switchly

11:17 as a solution architect.

11:21 Working with customers, what is it that,

11:22 how do you need our product to work?

11:25 What do we need to build additional capabilities

11:28 to make this work better for you?

11:29 Extremely bespoke, very custom to what they needed.

11:32 Very, very important, right?

11:35 Solving important problems that they had.

11:37 You're seeing extensions of it where it's like,

11:40 okay, honestly more of like solution architecture.

11:44 I would say the way that I think about solution architecture

11:47 where it's like, we've got our software

11:49 and I'm really just trying to help you use our software

11:52 pretty much out of the box, maybe extending some APIs,

11:54 but maybe writing some scripts, right?

11:57 But like helping make my technology work

12:01 in your environment, but not really doing any custom

12:03 development to get there, right?

12:05 Not really extending the product in that way.

12:08 I think we're seeing that theme emerge a lot.

12:11 We do that at Sourcegraph.

12:13 That is one of the core things

12:14 that our field engineering team does.

12:16 I know that's, you know, there's a big splash being made

12:18 with open AI and anthropic doing similar type of scope.

12:24 And so, you know, it's interesting to see kind of

12:27 the different iterations of this,

12:31 but I really feel like when you look at it

12:33 and you boil it down, the heart of what is happening

12:36 is its professional services, you know, for customers.

12:41 - What would be the difference?

12:42 What would be like, oh, this is a proper

12:45 forward deployed engineer.

12:46 They're using that term in the correct way

12:50 versus just throwing it around as a buzzword

12:52 and they're really doing pro serve.

12:54 What would be the line where you would say,

12:56 okay, yeah, that's what that term is for?

12:58 - Yeah, well, I mean, right, Palantir created it.

13:02 So I feel like their definition, you know,

13:05 at least the way that I think about it,

13:06 their definition is kind of the standard of that.

13:11 And I guess kind of two pieces, you know, that's separated.

13:18 One is, you know, are they actually paying?

13:21 Is it a paid engagement, right?

13:23 Is there any type of, you know, billable metric

13:29 to that customer for that work being done, of course,

13:33 would be kind of the first ones of that.

13:36 But I think the second one is, you know, in some cases,

13:39 you literally are just hiring the engineer.

13:43 It's kind of like engineer for rent.

13:45 I'm hiring you to come and like, and do this.

13:48 It could or could not have anything to do

13:51 with a product that they have.

13:54 I think that you see those elements

13:56 of like what I would call true for deployed engineer.

13:59 To me, if you're charging, to me,

14:02 if like you're using your tech,

14:03 if you're trying to embed your tech,

14:04 if you're trying to make your tech more stick,

14:06 if you're trying to make your tech work,

14:07 if you're trying to extend it right,

14:08 all of those things I would argue are professional services,

14:13 not for deployed engineering.

14:16 - Makes a ton of sense.

14:19 I was hoping we'd get into a little bit.

14:20 I think a lot of software companies struggle with,

14:25 okay, when do I need to bring in professional services?

14:29 And then how do I bake that into my model

14:31 and manage it without completely wrecking my gross margin?

14:36 What would give you a signal

14:39 that they need to start layering in

14:42 a proper professional services organization?

14:45 And how should they start building that out?

14:49 - Yeah, in my experience,

14:53 and I think that internally, culturally,

14:57 this also makes it so much easier

14:59 to stand that function up and integrate it,

15:04 our customers were always pulling us in that direction.

15:07 And so like an example of this at Sourcegraph,

15:10 when I first joined,

15:14 we did not have a cloud offering,

15:16 so all of our customers would install us on-prem,

15:19 so like in their own cloud service provider,

15:21 in their own data center.

15:23 And this wasn't something that we could,

15:27 that we were very good at helping them through.

15:30 It would take them a very long time to install,

15:33 it would take them a very long time

15:34 to get the hardware that they needed to get deployed,

15:38 configured, right, everything in place to maintain it.

15:41 And they were looking to us to help them,

15:44 like it was not uncommon at that point in time

15:47 for it to take six months.

15:49 And so when I say like,

15:51 customers were pulling us in this direction,

15:53 it's like asking us for that help,

15:55 saying that they need this,

15:59 that is so much easier to identify that need

16:03 and build that need and integrate that need

16:05 into your company,

16:06 rather than taking a separate approach of like,

16:10 listen, professional services can be extremely lucrative.

16:13 There are a lot of companies

16:15 that have done exceptionally well

16:17 and professional service has been

16:19 a very big revenue driver for them.

16:21 And so you could say, well, okay,

16:24 here's a potential revenue stream.

16:25 What services could we create?

16:27 What packages could we create?

16:28 And you can take that approach as well.

16:32 And so what's been really interesting,

16:35 even at Segment on the solution architecture side,

16:38 when people would buy Segment,

16:41 it's a Martech platform,

16:42 but you need to go and integrate Segment

16:44 into your website, onto your mobile apps, right?

16:48 There is definitely an implementation component of it.

16:51 And they were struggling.

16:52 And so similarly, they were like pulling us

16:54 in this direction of, hey, we need help.

16:57 And so identifying those patterns

17:00 and identifying those areas

17:01 where customers are actually like raising their hand

17:04 and identifying that like there is a true need

17:06 and you see that, right?

17:07 Emerging in different areas

17:10 is obviously the simplest flag

17:14 for you to know that there's an opportunity

17:17 or maybe you should consider

17:20 building out a dedicated function for that.

17:23 I think as well, this is certainly true for Sourcegraph.

17:28 The largest engineering organizations

17:30 are really the ones that get the most value from Sourcegraph.

17:33 These are the companies with tons of code,

17:35 multiple code hosts, et cetera, et cetera.

17:38 Sourcegraph provides a product capability

17:42 that just doesn't exist for them.

17:45 Those are also the types of companies

17:46 that are very used to working with vendors

17:49 who provide professional services.

17:51 And so there's also a little bit of that tug.

17:53 If you wanna play in the enterprise,

17:54 they're going to have certain expectations.

17:56 They're gonna expect to be able to pay

17:58 if they want for 24 by seven support.

18:00 They're going to expect that somebody can take over

18:04 the maintenance burden for them

18:05 and manage their upgrades if they so choose,

18:08 that they can ship a laptop to somebody

18:10 or give them VPN access.

18:11 These are all things that we experience.

18:13 And so it's like also very interesting how,

18:16 depending on who your product service is, right?

18:19 Who your ICP is, where you play,

18:22 you may also get pulled in that direction.

18:24 That was certainly true

18:25 for some of the professional services offerings

18:27 that we ended up building here at Sourcegraph.

18:31 - How do you draw the line between,

18:34 hey, maybe my product needs support

18:37 and maybe there's things we need to simplify in the product

18:40 or help the customer serve themselves versus,

18:44 no, this is actually a real opportunity

18:45 to help the customer and have a value-added service.

18:50 What are some things companies should be looking for

18:52 that would be signs that, hey, no,

18:54 this isn't a product problem where you need more support,

18:57 or it's not really a problem

18:59 that just support is designed to solve.

19:02 You really need to go a different direction.

19:04 What signals should they be looking for?

19:07 - It's really interesting that you asked that

19:08 because I can remember very deliberate conversations we had,

19:13 kind of going back to the example around helping customers

19:16 with implementation and deployment.

19:18 We had many very deliberate conversations.

19:22 Do we go and solve this in the product

19:24 or do we hire a team and centralize that expertise?

19:27 And I think the reality is there's no amount of product work

19:34 that could solve that problem 100% of the way,

19:38 different environments, different network configurations,

19:41 different needs, different tech, et cetera.

19:43 And so for us, we really did think about

19:49 that very deliberately.

19:51 And one of the things that we pride ourselves on

19:54 is being able to work in any customer environment.

19:57 Like that for us is like part of our differentiation.

20:01 And so if that is part of the core value prop

20:04 that we want to take forward,

20:07 if we wanna be able to make that statement

20:10 and stand behind it,

20:12 you have to be willing to do what it takes

20:15 to actually be able to follow through.

20:16 And for us, again, there's no amount of product improvements

20:21 that are going to make that possible 100% of the way.

20:26 And I think you said that it's like,

20:28 there's also a very proactive strategic opportunity

20:31 for us to get closer, get more embedded.

20:35 That there is absolutely like a very,

20:39 there's very clear data that tells us

20:41 that customers who partner with us,

20:42 who have our implementation services,

20:44 who update, who maintain their instance

20:48 and update regularly, et cetera, right?

20:50 Retain and grow at a higher clip than those that don't.

20:54 There's also like a lot of the data to back that up.

20:58 But you have to look at those types of considerations

21:04 or characteristics.

21:05 Could you truly go and solve this in the product?

21:08 I think most of the time, like you would start there.

21:12 But for us in some of these scenarios,

21:16 who we wanted to be and where we want to operate

21:19 and the types of customers we wanna work with,

21:23 again, we were like pulled in that direction of like,

21:25 we have to be able to bring that expertise

21:29 and bring that skill set to make source graph work

21:32 in any environment, any configuration.

21:36 - Well, let's lead with that then.

21:38 So let's assume customers are pulling you in,

21:41 they're asking for a lot of help.

21:42 It seems like you have an opportunity to help,

21:45 not just with the product,

21:46 but build out within context of their specific situation,

21:50 their unique tech stack, industry,

21:53 firmographic design and setup.

21:55 And you're about to build this department.

21:58 I think the biggest things tech companies

22:02 would probably run into is,

22:05 how do I price and package this?

22:08 And what benchmarks should I be looking at

22:12 in terms of staffing capacity, gross margin?

22:16 How many people should I have?

22:17 And how much revenue should a pro serve org be bringing in

22:21 at least relative to the spend?

22:22 - Yeah, it's definitely a maturity curve, right?

22:26 You have to start somewhere,

22:32 you have to make the decision

22:34 and commit to we're gonna build this team out.

22:38 We want to be able to say yes to this capability, right?

22:42 So you kind of start there.

22:43 There was a lot that we had to learn.

22:45 And the same was true at Segment as well.

22:48 The first real task for us

22:51 was to figure out how to do

22:52 these implementations repeatedly.

22:56 We really needed to be able to confidently

23:00 close a new customer and say,

23:02 okay, if they follow this playbook,

23:04 they should be up and running in 30 days, right?

23:08 However many weeks or days that project was expected to be.

23:13 And so you kind of have to start there.

23:16 And you start by learning how to do it, right?

23:18 Bringing people with, hiring people with the right skills

23:21 and capabilities to start to figure out how to do this,

23:24 build the playbook,

23:25 how do we start to be able to do the same thing

23:28 within the same general timeframe.

23:30 Once you start to build out a little bit of that competency,

23:34 then you can start to think about,

23:36 okay, what does the packaging of this look like?

23:39 Because you're thinking about the packaging,

23:41 there's the head count, there's the pricing of it, right?

23:43 But there's also, you're committing to deliverables,

23:46 you're committing to timelines,

23:47 you're committing to these outcomes that you're delivering.

23:51 And that's an important piece of when you think about

23:56 inserting that into the sales conversations,

23:58 when you're talking about, what is this line item?

24:01 Why should I buy it?

24:02 You need to have kind of that level of detail and specificity

24:06 to be able to paint a very clear picture for them

24:09 of what this is, what this looks like,

24:11 what you can expect from us.

24:13 You don't get there from the beginning.

24:15 You get there from deciding

24:17 that you're going to step into this.

24:19 And so we knew that we needed to reduce the time to value.

24:25 We knew we needed to figure out how to help customers

24:27 very consistently get up and running quickly.

24:30 We hired incredible experts that are very seasoned

24:36 with Kubernetes, with all of the different

24:39 types of configurations that our customers might run into.

24:45 And we started doing it, we started learning

24:47 and we started pulling the playbooks out of that.

24:50 And once we had that, we said,

24:51 okay, what does this look like

24:53 for us to actually go and sell it?

24:56 And as well, right, some of that

24:59 is a little bit of a guessing game.

25:00 There's lots of data out there, right?

25:03 Talk to other companies, VCs can share some of that data

25:08 as well, certainly, from my past experience,

25:11 had some data points.

25:13 So there's certainly a starting point

25:15 that you're thinking about.

25:17 But some of it is like,

25:18 you start to have these early conversations

25:20 and the market is going to adjust you

25:23 pretty quickly on what they think is reasonable.

25:26 And you have to really think about

25:28 the value of the package.

25:31 For us, there are a couple services

25:33 that are mandatory attach.

25:35 So if a customer self hosts,

25:37 mandatory, you have to work with our implementation team

25:42 because we know that it leads to better outcomes.

25:44 We know that time to value.

25:46 It's priced at a really, really low reasonable point

25:50 to avoid any friction.

25:53 But there are some services that we have

25:57 that I would say resemble more of what you see

26:01 from forward deployed engineering

26:04 in the way that we're thinking about it.

26:06 And for that, you really have to help them

26:09 have a clear picture of like,

26:12 what is the justification behind this price?

26:14 Well, we're doing all of these things

26:15 and here's what you can expect

26:16 and here's what we deliver and da, da, da, da, da.

26:18 Right, you have to have the playbook

26:20 to really make that clear for them.

26:22 It's like an alignment of value to the services provided.

26:26 And so it's like, that's kind of like,

26:28 this like second step.

26:29 And then it takes time for you to get to the place

26:33 where you're running it as a P&L,

26:35 you've got the utilization targets,

26:37 you're thinking about kind of the margins behind it.

26:41 Like you get there, I would say like,

26:44 that's the third or fourth step in line.

26:49 What are some ways that you price and package?

26:52 Are you doing packages of hours?

26:54 Are you tying it to projects, fractional access to people?

27:00 Does it depend on what you're doing?

27:02 What are some packaging models that you've seen work well?

27:05 Yeah, absolutely depends on what you're doing.

27:08 Like I mentioned, so mandatory attach,

27:11 when customers self-host, we have like the base 10K

27:15 and then we have the advanced 25K.

27:18 And so what that gets you,

27:21 it's essentially a series of activities

27:23 and things that we're doing.

27:24 We're helping you install, we're helping you deploy.

27:27 We're helping you monthly update your instance.

27:31 We're helping you quarterly do maintenance, right?

27:35 Making sure there's not anything that we need to fine tune

27:38 to keep the instance performant, et cetera.

27:40 And so, do you have one instance?

27:43 Do you have multiple instances, right?

27:44 That kind of helps us understand,

27:46 should you be in the 10K package or the 25K package?

27:48 But it's that set fee and you're getting a sense of like

27:52 throughout that annual contract, the different services

27:55 and the touch points that you're getting with the team.

27:57 So, right, that's fairly standard.

27:59 Similarly, that was kind of the base package

28:03 at segment as well.

28:05 You know, they paid that fixed amount.

28:07 We would come in, we would help them to install, right?

28:10 Get the data pipeline up and running,

28:12 get their integrations,

28:14 their downstream destinations configured at six week,

28:17 you know, six week, 12 week project, and then you're done.

28:20 So it was like kind of that fixed cost.

28:22 We've also done it, we have a resident architect package

28:27 where, you know, we have customers who,

28:30 they want things done in a certain way.

28:32 They want onsite trainings.

28:34 They want materials that use their terms

28:38 or use their language or, right?

28:40 Highly, highly customized to their environment.

28:42 And so in those situations, we do sell packages of ours.

28:47 And there was a similar model as well of that at segment two.

28:51 But, you know, what are the types of things you want to do?

28:55 What, you know, work streams are we doing train the trainer?

28:58 Are we coming onsite once a quarter, right?

29:00 Like let's get a sense of what those activities

29:02 are that you want.

29:04 And then you're essentially buying a bucket of hours

29:06 for that year that you burn down.

29:09 And, you know, it works out to avoid, you know,

29:12 it's not feasible to say, oh, we have 200 hours left

29:16 and our contract expires in a week.

29:17 We want to use those, right?

29:19 You net it out to, you know, a weekly breakdown.

29:23 Maybe there's a work stream and, you know, we pull from that.

29:26 But in general, they're buying, you know, a bucket of hours.

29:31 So there's like, you know, that's the second flavor of it.

29:34 The third flavor that we have is,

29:38 and, you know, we were doing this more.

29:40 So Sourcegraph, we split off from our coding agent,

29:45 the coding agent part of our company, AMP.

29:47 And AMP was really doing forward deployed engineer.

29:51 Like you're, you know, you're paying 500k, you know,

29:56 whatever the dollar amount may be,

29:57 you're paying some amount of money

30:00 and you're getting access to these engineers

30:02 that are going to come in

30:03 and we're going to deliver an outcome.

30:06 We're going to help you with this project.

30:10 We're going to help you with this migration.

30:11 We're going to help you activate this workflow, right?

30:14 And so, and, you know,

30:16 maybe we're doing five of those projects for you, right?

30:19 And so it's like more tied to, we're hearing,

30:22 we're hearing this of this emerging theme

30:24 around outcome based pricing.

30:25 But I would say like that FDE type package

30:28 is more tied to like an outcome that's being delivered.

30:31 And we're agreeing to some dollar amount that, you know,

30:35 we attach to that.

30:38 - And as it sounds like at the beginning,

30:40 maybe you're operating this at break even,

30:43 maybe even a loss just to get the traction

30:46 with your product.

30:47 As it matures, what should you be working towards

30:52 in terms of utilization and gross margin of the overall org?

30:57 Let's assume it'll probably take some steps to get there.

31:00 But when you're kind of running this at full maturity,

31:02 what does that tend to look like?

31:04 - Yeah.

31:05 Yeah, I mean, you know, when you get to full maturity,

31:08 you have a sense of, you've done enough of these

31:13 that there's like predictability in the pipeline, right?

31:16 So you can correlate and, you know,

31:18 in the same way in sales, we're planning ahead, right?

31:21 Our quarters, our years, you can look at that

31:23 and you can plan out, okay, I should expect

31:26 this many types of engagements, this much money coming in.

31:29 Here's, you know, what my team can handle, you know,

31:33 is it, we're looking for 80% utilization,

31:35 90% utilization, right?

31:37 What is that that we want them billable?

31:39 And you're, you know, you're forecasting that growth,

31:41 you're forecasting a revenue target,

31:43 you're forecasting head count, right?

31:44 And you're able to have a lot of predictability.

31:48 - You're almost running a company

31:50 within the company at that point.

31:52 - Yeah, absolutely.

31:53 I mean, and I saw this firsthand,

31:55 like when I was at IBM, I had a utilization target.

31:58 I was, you were not expected to be, you know,

32:00 on the bench, you were expected to be

32:01 staffed on projects, delivering for customers, right?

32:05 And so, you know, early on in my career,

32:10 this was something that I myself was,

32:13 I was, I was metric and gold on.

32:16 And so you get, you know, you get,

32:18 you get to that level of maturity.

32:22 You know, one thing that I'm always thinking about is,

32:25 you know, you want to get to that breakeven point.

32:27 Like you said, you, you started a loss, right?

32:29 You kind of have, I feel like you kind of have to,

32:31 you need to really learn--

32:33 - Just to start getting traction

32:34 and know what you're going to do, right?

32:35 - Yeah, exactly.

32:36 So then, you know, the next step you're trying to,

32:39 you're trying to break even from that.

32:42 And, you know, you're getting a sense of,

32:44 okay, well, what's the fully loaded cost

32:46 of like this employee of this team?

32:47 How much can this team handle?

32:49 You know, that's like, that's a really good exercise

32:52 for validating some of your packaging.

32:54 It's like a dimension of it.

32:56 Again, the market, the customer, you know,

32:59 customers are going to correct you.

33:01 If, if they don't feel that your services align, right?

33:06 The value aligns to the sticker price.

33:10 But another kind of useful data point,

33:13 going back to the example of these implementations

33:15 that we do, it's like, you know,

33:17 what's the fully loaded cost of,

33:19 of one of the engineers on this team?

33:21 Okay, how many of these can they do at once?

33:23 Okay, you know, what would that,

33:25 what would that price point need to look like

33:28 for this to make sense for us to even break even?

33:30 Okay, well, what does that look like

33:31 if we want to get a million dollars

33:34 of professional services revenue by the end of the year?

33:36 Right, what does that look like?

33:37 And so those types of data points and conversations

33:40 are also really helpful when you're thinking about

33:43 how to price this, what does it entail?

33:46 How much should we be doing

33:49 for any one of these engagements?

33:50 How many people do I need, et cetera?

33:54 - Should companies be looking at this as,

33:58 hey, this is something that helps get the adoption

34:01 of my product and kind of running near break even is okay

34:06 or should they be looking at this as a decent profit center

34:10 that they can drive real growth from?

34:15 - I've seen both, I've seen both for sure.

34:19 I think some of that does somewhat depend on,

34:23 a bit on the DNA of the company.

34:26 And I think, I think, and you know, I said this earlier

34:29 but I also think some of that depends

34:31 on where you're operating.

34:33 Like for us, you know, for us operating in the large

34:36 enterprise, professional services are not a one-off thing.

34:42 You know, like they work with tons of vendors.

34:45 They, you know, in some cases just expect to have this.

34:49 I think when you're operating in those types of environments

34:52 with a really extensible type of product,

34:56 a very complicated yet powerful type product,

35:00 you have a real opportunity to turn that

35:02 into a growth lover for your business.

35:06 But, you know, if you're just looking at it as like,

35:10 oh, here's a potential revenue stream.

35:12 Let me go try to make this, you know,

35:14 square peg fit in a round hole.

35:17 That's gonna be really difficult culturally, right?

35:20 Internally, and that's going to require a lot more,

35:23 I think, to make successful.

35:25 It's one of the reasons why I love that, you know,

35:28 where I've found myself in companies really building

35:32 and using professional services,

35:35 the customers are pulling us in that direction.

35:40 Makes a ton of sense.

35:41 One thing I see too, I've noticed this

35:44 in the go-to-market tech space quite a bit.

35:47 So if you look at companies like Salesforce, HubSpot,

35:50 Clay, they tend to outsource a lot of the services

35:54 to partners.

35:57 And then some companies say, no,

35:59 we're gonna own the full experience.

36:02 On the spectrum to fully outsourcing,

36:04 to fully having in-house,

36:07 what are some of the reasons to go one direction

36:09 or the other if somebody's not sure

36:11 how to sort through that?

36:12 - Yeah.

36:13 I've encountered that as well, even with vendors,

36:16 you know, across tools in our revenue stack.

36:20 Mandatory services, maybe you have to go

36:22 with this outsourced third party.

36:25 From my experience, having, there is such value

36:29 in having that competency in-house.

36:32 So, you know, my experience, what I've seen,

36:36 I've lived through it as an IC,

36:37 I've lived through it now multiple times, right?

36:39 As a leader, there is such value in having

36:43 at least a portion of that experience in-house.

36:46 You know, you're going, you're learning things,

36:47 you're connecting with customers, right?

36:52 You're getting more deeply ingrained.

36:54 You're hearing feedback that leads to improving

36:57 the product in some places.

36:58 You're, you know, going and building something custom

37:01 and you get this idea to go productize that.

37:03 That feedback loop, I have not seen that done well

37:07 when you're strictly using outsourced,

37:09 you know, outsourced agencies.

37:12 I think that you can start to outsource,

37:15 you know, it's super, super helpful

37:16 for like regional type support.

37:19 At Segment, you know, we had multiple vendors

37:23 in different parts of the world to really help us,

37:27 to help us give a very consistent experience globally.

37:32 But, you know, there were absolutely some projects

37:35 that due to the complexity or due to the strategic nature

37:39 of the logo or the, you know,

37:42 what they were trying to accomplish

37:43 that we absolutely wanted our internal teams attached

37:46 to that so that we could learn,

37:48 so that we could strengthen the relationship,

37:50 so that we could use those learnings to go and, you know,

37:52 build out that industry vertical, make our product better.

37:56 So I think that there are very strategic reasons

37:58 for taking that on.

38:00 And then as well, you know, you outsource it.

38:03 There is an overhead to that, right?

38:07 Somebody has to manage those relationships.

38:10 Somebody has to make sure that they are delivering

38:13 the service at the quality that you would expect.

38:16 You know, it's your brand that you're outsourcing

38:19 and having them take forth.

38:20 And so how do you keep them,

38:22 how do you make sure that they get enabled and up to speed?

38:25 How do you make sure that they know, you know,

38:27 how the product is evolving?

38:29 How do you make sure that they're able to provide,

38:31 you know, the level of support

38:33 that you would provide, et cetera?

38:34 And so there's also that very real consideration.

38:37 It can seem very appealing of like, yeah, let's just,

38:39 you know, let's just outsource this, but you have to know

38:42 and you have to be ready to solve

38:45 for those overhead dynamics.

38:50 - I've definitely seen it done wrong

38:52 where a company has completely outsourced

38:57 maybe things like implementation.

38:59 So like, no, we want pure tech revenue.

39:00 It's got the highest multiple.

39:02 Maybe that revenue is gonna dilute.

39:05 And it's like, maybe, but probably what's more damaging

39:09 is a partner not implementing your technology very well.

39:13 And you're not investing enough to manage them.

39:16 As you mentioned, it's not for free.

39:18 It's just another strategy.

39:21 The only time I've seen where maybe it starts to make sense

39:25 is if there's a certain level of industry

39:27 or domain expertise where maybe, hey,

39:30 this is more of a horizontal type of tech play.

39:35 And we have partners that are specialized

39:38 in certain industries, but if you're just expecting people

39:41 to know your product better than you are,

39:43 it seems like a risky bet to me.

39:45 - Yep, yep.

39:46 And you know, you might get to,

39:48 you're trying to do too many things, right?

39:50 It's easier to outsource that.

39:51 Like you're growing really fast.

39:53 You have more customers you can't hire.

39:56 It can seem very, very appealing, truly.

40:00 But yeah, you're spot on.

40:03 Like a customer buys your software,

40:07 they wanna get it up and running.

40:09 And if they are having problems,

40:11 if they're not getting what they need,

40:14 they're not thinking about that third party.

40:16 They're thinking about, man,

40:17 the software doesn't work for me.

40:19 This isn't delivering what I hoped it would.

40:22 And so, right, there's that very real risk to it.

40:26 - Yeah, you're gonna end up getting the blame

40:28 if you don't fully own the experience.

40:32 I was wondering, you know,

40:34 'cause I've run a RevOps agency,

40:36 so I'm helping fast companies,

40:40 B2B SaaS companies grow in scale

40:42 and help them augment the RevOps teams.

40:45 This is an interesting area where I feel

40:47 like RevOps doesn't talk too much about.

40:50 Do you lean on RevOps to help with things like forecasting

40:54 for your professional services,

40:57 helping them with their internal tech stack and tools?

41:00 It's a department that a lot of RevOps people

41:03 are not talking about helping.

41:04 - Yeah.

41:08 Yes and no, I would say.

41:11 Partly, I have an incredible PS leader right now

41:16 who really, you know, you talk about,

41:18 you think of it as like your own business within,

41:21 in the business.

41:22 He really has that level of ownership over it.

41:25 You know, we think about and, you know,

41:28 the available bandwidth and what do we need

41:31 and what's, again, the fully loaded cost, right?

41:33 He's thinking about those elements

41:37 of his team, of the performance.

41:41 We've been fortunate enough to, you know,

41:44 I mentioned kind of, there are packages

41:46 where we sell buckets of hours, right?

41:48 Being able to track that.

41:50 There are some great extensions off of Salesforce

41:53 that exist so that you can do that right

41:55 from within the one system.

41:58 So it's already the tool in our stack.

42:00 We haven't needed to, you know, necessarily go buy a new one,

42:04 able to report, right, in our system of record, et cetera.

42:08 So it's like, we've been fortunate to,

42:11 where we've needed to do that,

42:12 find tools, find plugins, you know,

42:15 connectors within those existing tools.

42:20 I've leaned on a lot our finance,

42:23 our finance team of helping us do some of the analysis

42:26 around this over the past several years.

42:30 But RevOps, you know, certainly overseeing our tooling,

42:34 overseeing the tech stack, making sure that, you know,

42:37 we're reporting on what we need to report on,

42:40 you know, et cetera, et cetera.

42:41 Like they are definitely a key part of that for sure.

42:46 - Yeah, I was just wondering,

42:47 I feel like it's kind of a big miss

42:48 for some RevOps teams to lean in and offer some help.

42:54 It's fantastic you have the leader that you have,

42:55 but I imagine a lot of companies are operating

42:59 without maybe the analytics firepower or technical,

43:04 probably very technical on whatever platform

43:07 they're supporting,

43:08 but technical on the productivity stack

43:10 that the team's using.

43:12 Just wondering if it would make sense

43:13 for those teams to lean in

43:16 and then help in that area of the business too.

43:18 - Yeah, I mean, yeah, absolutely.

43:21 Like I mentioned, you know, you're thinking,

43:23 you get to a point where it's more common than not

43:27 that somebody that, you know,

43:29 you bring on a new customer or even existing customers,

43:32 right, they want to leverage these services.

43:36 You know, that becomes part of the forecasting

43:39 and part of the planning.

43:40 And you're thinking about, right, that potential stream

43:43 and like RevOps obviously plays a huge part in that.

43:47 So there is a lot of data analysis that can be done

43:51 looking at different characteristics, right?

43:53 Looking at, you know, this type of customer,

43:57 they, you know, they bought these services,

43:59 they use that, right?

44:02 It's definitely a piece

44:04 that needs to come into your annual planning.

44:09 It needs to be considered.

44:11 And, you know, my professional services leader,

44:13 he, you know, he has a revenue target.

44:15 He has a services revenue target tied to his variable comp.

44:19 Like we want him, you know, we want him and the team

44:22 thinking about that, thinking very strategically,

44:24 thinking very proactively, getting integrated into deals,

44:29 figuring out where there's, you know, early risk.

44:31 Where's a customer trying to do something

44:33 that is, you know, really not by the book

44:36 or not a supported what we would call deployment method.

44:39 Let's identify that early.

44:40 Let's get into those conversations.

44:42 And so it's also like helping us be a lot more proactive

44:45 at some of this identification as a company.

44:51 - As a new CRO, where are you going to stay sharp

44:56 and to learn?

44:59 - Well, I am very lucky to have many mentors to lean on.

45:04 That first and foremost, I listened to a lot of podcasts.

45:13 You know, lots of podcasts, I'm always very hungry.

45:16 What are other people doing?

45:18 I try to have connections with folks in my network,

45:23 you know, meet on a quarterly basis, monthly basis,

45:25 try to trade notes.

45:26 What are you doing?

45:27 What are you seeing?

45:28 Here's a problem I'm having, you know, what would you do?

45:32 And so, you know, there's a lot of information out there,

45:38 you know, for folks to learn,

45:40 to hear what people are doing.

45:44 But yeah, I find, and you know,

45:47 I think part of this is like not being behind the computer.

45:50 I like to network, I like to connect, I like to learn.

45:53 And, you know, this whole topic of forward deployed engineer,

45:56 I've had a lot of conversations with people thinking about

45:59 and trying to build this and sharing my notes

46:02 of like what I've seen and what I would do different,

46:04 you know, what I would do differently

46:05 if I was starting this over,

46:06 as they're starting to step into that journey.

46:09 And there's just so much that you can learn very quickly

46:14 from others who have gone through this

46:16 or are trying to figure this out and write your trading notes

46:19 and kind of getting to that place together.

46:21 And like, also listen, I have a fantastic team, right?

46:26 You know, supporting me.

46:27 I have a wonderful, you know, global head of sales

46:30 that I learn from every day.

46:33 And there's a great, you know, collaboration between that.

46:36 And so there's definitely been a learning curve.

46:40 You know, I'd be lying if I didn't say

46:42 there was some imposter syndrome early on,

46:45 but using my network, using my mentors,

46:48 there's great articles, there's books out there, right?

46:52 You know, podcasts like this.

46:53 There's so much that you can learn

46:55 and adapt about yourself.

46:58 - Yeah, I think if you're not feeling

46:59 a touch of imposter syndrome,

47:00 you're probably not challenging yourself enough.

47:02 So I always look at that as a good sign.

47:05 If it feels like I should be in the room,

47:06 then maybe I'm not pushing hard enough.

47:09 But no, I really appreciate that.

47:12 I just think it's, things are changing so quickly.

47:16 And it's really, really hard to keep up

47:19 and then sift through the noise and find the signal of,

47:22 hey, what are people actually doing?

47:24 How are people actually operating?

47:26 And that's what we try to make this whole podcast about too,

47:28 is not just launching buzzwords,

47:32 not just launching strategic things,

47:36 but what are you actually doing?

47:38 What's the execution layer behind

47:40 how you're getting this stuff done?

47:43 And I appreciate all the details that you have shared

47:47 about this part of the business because,

47:50 especially with AI, a lot of the product moat

47:55 is starting to shrink.

47:58 And a lot of the competitive advantage

47:59 is coming on the value added service side of the business.

48:03 A lot of service, or sorry, a lot of software companies

48:06 are now starting to look a little bit more

48:07 like service companies.

48:09 And there's just an entire industry

48:11 that's never even really thought about doing that.

48:13 That's gonna need to start implementing these teams,

48:16 need to start managing this whole new type of business

48:19 that they haven't had to manage before.

48:22 And it's happening really quickly.

48:23 So I think a lot of the frameworks and roadmap

48:27 that you laid out will be really helpful

48:30 for anybody that's going through that transition right now.

48:32 - Yeah, that's great.

48:34 You just mentioned that we're, through Q2,

48:36 we've been really building out

48:39 our command of the message framework.

48:40 We're really trying to get oriented

48:42 around value-based delivery.

48:44 And one of our core differentiators is,

48:51 we're framing it as adoption-led delivery,

48:53 but it is absolutely a competency for us.

48:56 We've got years and years of experience of doing this.

48:59 We've built out those playbooks.

49:00 We can do it very consistently and repeatedly.

49:03 And it is absolutely a differentiator.

49:06 And to your point, you're seeing this come about

49:10 with a lot of companies.

49:11 And so I think that maturity curve is very real.

49:16 Learn to do it well, learn to do it consistently first,

49:21 and then you kind of earn the right

49:22 to start to charge for it.

49:25 But it is such a core part of the value

49:30 that we unlock for customers

49:32 and the value that they get from our platform.

49:33 It's based off of the people and the experts

49:37 and the domain expertise that they bring

49:39 to those engagements.

49:41 - Absolutely.

49:43 I think everything you're doing

49:45 is really exemplifying the trend

49:48 that the market is going in,

49:50 especially with software companies

49:52 needing to differentiate with their services.

49:53 So I think everything you mentioned,

49:56 it's a huge competitive advantage.

49:57 And people really need to get the value

50:00 out of the technology that they're buying

50:02 as quickly as possible.

50:03 So Amy, I just want to say thank you

50:06 for being on the podcast, being on the show,

50:08 sharing everything you did, going into detail,

50:10 the nitty and the gritty of it.

50:12 And I think it's such a cool thing

50:15 to see a different path to Chief Revenue Officer as well.

50:20 So a lot of people still have in their minds,

50:22 like, oh, I got to go.

50:24 S-D-R-A-E sales manager, that's the only path to CRO.

50:28 But I think we're seeing a lot of these operational routes

50:33 to get to the highest seat in the revenue org.

50:36 And it's excellent to have an example of that on the show.

50:39 So thank you, Amy.

50:40 I appreciate everything you have shared.

50:43 And I can't wait to see what you and Sourcecraft do next.

50:46 - Awesome.

50:47 Thanks for having me.

50:48 Enjoy the chat.